I fix hard business problems through operations, technology, product partnership, and organizational design — turning fragile revenue into durable growth.
Brands I've worked with
About
For the past decade I've worked at the intersection of customer success, commercial strategy, and organizational design. Most recently as Head of Platform Growth & Performance at Wunderkind, I built a dedicated business unit from scratch, led a global platform migration, and redesigned the operating model around it.
Before Wunderkind, I scaled post-sale operations across 16 M&A-integrated companies at Recurrent Media and led brand partnerships at Task & Purpose, closing multi-six-figure campaigns with Nike, Under Armour, Ford, and others.
I'm currently building AI-native workflows into the core of how customer-facing organizations operate. Not as a side experiment — as the foundation for how teams scale without adding headcount.
I served six years in the U.S. Marine Corps and sit on the Board of Advisors for The Headstrong Project, a national nonprofit providing free mental health treatment to post-9/11 veterans.
"Not as a side experiment — as the foundation for how teams scale without adding headcount."
Case Studies
64% of churning clients shared a single root cause. I found it by analyzing Gong sentiment and Salesforce data with AI, wrote the diagnosis memo to leadership, then designed and executed the migration framework that became Wunderkind's company-wide strategy.
Read Case Study →Before most CS organizations were thinking about AI, I designed and deployed a suite of practical tools that transformed how the team operated. Tasks that took hours dropped to five minutes. Capacity scaled without headcount.
Read Case Study →Renewals were fragmented, forecasting was unreliable, account ownership was unclear. I redesigned the operating model from the ground up — removing the Client Partner layer, giving CSMs full commercial ownership, and rebuilding renewal governance from scratch.
Read Case Study →Writing
Coming Soon
CS teams spend most of their time reacting to problems that were predictable weeks earlier. Here's the framework — built on AI, health scoring, and scalable playbooks — that changes that equation.
→Coming Soon
Layered account teams feel like coverage. They're actually diffusion of accountability. The case for giving CSMs full commercial ownership — and how to make the transition without losing revenue.
→Jun 2026
Agents operate in a black box. Troubleshooting burns hours and dollars. Here's what I found when I stripped it back.
Read →Resume
Head of Platform Growth & Performance
Wunderkind
Built and led a dedicated business unit — product alignment, post-sale org design, commercial strategy, and margin improvement.
Director, Customer Success
Wunderkind
Managed $25M+ mid-market portfolio. 98% gross retention, 120% NRR. Built CS processes, automation tooling, and account planning infrastructure.
Head of Customer Success
Recurrent Media
Scaled post-sale operations across 16 companies integrated through M&A. Built unified processes from contract through renewal.
Director, Partnerships & Video Strategy
Task & Purpose
Closed multi-six-figure brand partnerships. Grew YouTube from 12K to 240K subscribers in 12 months.
Operations Manager & Account Executive
Ranger Up
Built B2B account management processes and digital marketing operations for an e-commerce and media brand.
Service
Board of Advisors
The Headstrong Project
National non-profit providing free mental health treatment to post-9/11 veterans and their families.
Current
Veteran
U.S. Marine Corps
2005 – 2011
These case studies contain proprietary details. Enter the password to continue.
Incorrect password. Try again.
Cancel64% of churning clients shared a single root cause that wasn't being addressed. I found it, named it, and built the strategy that became Wunderkind's company-wide answer to its most existential competitive threat.
Chapter 01
Q1 2025 — finding the real problem
The internal story was familiar: price sensitivity, rising competition — Klaviyo, Attentive, Retention.com all moving into the space. Reasonable. Also wrong.
I decided to look at the data differently. Using AI to analyze sentiment patterns across Gong call recordings and Salesforce notes, I reviewed churn accounts across every cohort — mid-market, enterprise, and strategic.
"64% of churning clients shared a root cause that had nothing to do with price or competitive features. The theme was control."
Clients described Wunderkind's managed-service model as a "black box." They couldn't see what was happening inside their own programs, couldn't make changes without waiting weeks, didn't own their data, and couldn't integrate cleanly with the tools they were already running.
Klaviyo wasn't winning because it was cheaper or technically superior. It was winning because it gave marketers autonomy.
I cleaned up the findings and wrote a briefing memo to leadership: this wasn't a competitive problem. It was a product-model problem. The solution wasn't to defend the managed service — it was to evolve beyond it.
Chapter 02
The strategic recommendation
The memo proposed a shift in positioning. Instead of treating Klaviyo as a rival to displace, Wunderkind should lean into integration — building its identity resolution layer to operate natively inside the ESPs clients were already using.
"The pitch to clients wasn't 'switch platforms.' It was 'stay in Klaviyo, add Wunderkind's intelligence underneath.'"
This became Signals — a product that passed Wunderkind's six behavioral triggers (abandoned cart, product, category, back-in-stock, low-in-stock, price drop) directly into Klaviyo flows. Clients retained creative control, owned their templates, and managed their own journeys — while still benefiting from Wunderkind's identity graph at scale.
The structural outcome was the Klaviyo App Marketplace partnership: Wunderkind listed as a Preferred Tech Partner, officially moving from perceived competitor to integration layer. This is now Wunderkind's company-wide direction, extended across the entire ESP and CRM ecosystem.
Chapter 03
Building the migration infrastructure from scratch
There was no existing framework for moving 83 active clients off a managed-service model and onto a self-service integration without losing them. I designed a three-phase approach with defined owners, dependencies, and exit criteria at each stage.
Phase 01
Scope & Prep
Audit abandonment flows, coupon delivery logic, and on-site dependencies. Identify red flags per account before any migration proposal.
Phase 02
Commercial
Pitch Signals to each client, secure signed work orders, update Salesforce opportunities to maintain ARR alignment.
Phase 03
Migration
Duplicate backend flows, build native Klaviyo flows, complete QA with ProServ, and execute go-live.
I built and delivered a CS enablement training program covering ticket submission, account scoping, ProServ routing, and QA criteria. I also pushed Product to ship automation tooling that compressed the migration from 3–4 weeks to 4–6 business days per account.
I personally owned 26+ accounts through the migration — including Bissell ($391K ARR) and BrainMD ($326K ARR) — managing commercial conversations, scoping, and go-live coordination directly.
Chapter 04
Before & after
Before
57.1%
Renewal rate — declining quarterly
After
98%
Gross retention · $19M ARR Beta scale · 83 clients migrated
The Klaviyo migration became the blueprint for Wunderkind's partner-first product direction across the entire ESP and CRM ecosystem. The insight from the churn analysis — that clients wanted to own their programs, not outsource them — reshaped how the company thinks about product development, packaging, and go-to-market.
Wunderkind × Klaviyo — official partnership announcement
Official Announcement
The Klaviyo App Marketplace partnership, officially in market.
Read on Wunderkind.co →Before most CS organizations were thinking about AI, I designed and deployed a suite of practical tools that transformed how the team operated. Tasks that took hours dropped to five minutes. Capacity scaled without headcount.
Chapter 01
The operational bottleneck
Customer Success teams at scale face a structural problem: the work that builds relationships — strategy, insight, proactive outreach — gets crowded out by the work that just has to happen — reporting, data prep, product configuration.
In 2023, I mapped exactly where time was going. The majority of hours spent on reporting, product grid creation, and CSV processing were candidates for automation — not because the work wasn't important, but because it didn't require a human to do it from scratch every time.
"The goal wasn't to replace the team. It was to give them back the time to do the work that actually required them."
Chapter 02
What I built
I designed and deployed three core tools that went into daily use across the CS organization:
Product Grid Creation Tool — Automated the assembly of client-specific product grids, pulling from data inputs and formatting for delivery. Reduced prep time from 2+ hours to under 10 minutes.
Signals Reporting Builder — Generated standardized performance reports for Signals clients, pulling metrics and formatting narratives automatically. Eliminated a weekly recurring bottleneck.
CSV Ingestion & Chart Generation — Automated data processing and visualization for client-facing deliverables. What previously required manual spreadsheet work now ran in minutes.
Chapter 03
What changed
Before
2–4h
Per reporting or grid task
After
5 min
67% faster time-to-value for clients
The team didn't grow to absorb the Signals migration workload. The tools did. The operating model scaled because the repeatable work was automated, not because we added headcount to absorb it.
This approach — building the AI layer into the workflow rather than bolting it on — became the foundation for how I think about CS operations at scale.
Renewals were fragmented, forecasting was unreliable, and account ownership was unclear. I redesigned the operating model from the ground up — and hit 120% NRR.
Chapter 01
The structural problem
When I took over the mid-market CS organization in 2022, the team had a structural problem that looked like a performance problem. Renewals were fragmented. Forecasting was unreliable. No one could clearly say who owned what outcome on any given account.
The root cause was the Client Partner model — a layer of relationship managers sitting between CSMs and commercial outcomes. In theory, it provided specialization. In practice, it split accountability, diffused ownership, and created handoff gaps that clients felt.
"Accountability that belongs to everyone belongs to no one. The first change was making that explicit."
Chapter 02
What I changed
Eliminated the Client Partner layer. Removed the intermediate role that was creating accountability diffusion. CSMs became the primary commercial relationship — responsible for renewal, expansion, and escalation.
Introduced Portfolio Managers. Senior ICs with broader account scope and a clearer mandate for strategic growth. Distinct from the old CP model because they held direct revenue accountability.
Built new governance frameworks. Forecasting, escalation, and renewal processes rebuilt from scratch with clear owners, defined cadences, and measurable outputs at each stage.
The transition wasn't frictionless. Moving CSMs into full commercial ownership required coaching, tooling changes, and a period where both models ran in parallel. The short-term disruption was real. The long-term clarity was worth it.
Chapter 03
The outcome
Before
Fragmented
Split accountability, unreliable forecasting, unclear ownership
After
120% NRR
48% margin improvement · $25M+ portfolio managed
The redesign produced results across every metric that mattered. Net revenue retention hit 120%. Margin improved 48%. Forecasting accuracy — previously a persistent problem — became reliable enough to support executive planning cycles.
The model — CSM commercial ownership, Portfolio Manager scope, clear escalation governance — became the template for how the organization scaled from there.
Agents operate in a black box. Troubleshooting burns hours and dollars. Here's what I found when I stripped it back.
There's a thread making the rounds on Reddit: "Is markdown and folders all we need now?" The comments are predictably split — engineers defending their agentic frameworks, practitioners quietly admitting they keep gravitating back to a plaintext file and a sensible directory structure.
I've built real AI workflows inside a business — not demos, not side projects, but tools a Customer Success team of dozens used every day. And I keep landing in the same place: the biggest advantage of markdown and folders isn't that it forces you to get organized. It's that it gives you and the AI a shared interface you can both inspect.
When you build an agentic workflow, the "thinking" happens inside a context window and disappears. State lives in memory layers you didn't design. Tool calls fire in sequences you can't easily follow. When something goes wrong — and it will — debugging means reverse-engineering what the agent decided, not reading a file.
You're not collaborating with the AI. You're delegating to a black box and auditing the output after the fact.
I've watched teams spend weeks building orchestration frameworks on top of workflows that were fundamentally broken at the information layer. The agent wasn't failing because the architecture was wrong. It was failing because the knowledge it needed was scattered, implicit, and unwritten. No amount of tool-calling fixes that.
Here's what changes when you work with markdown files and folders: the AI reads what you wrote, writes back to the same place, and you can inspect every step. The file system becomes a contract between you and the model — one that's human-readable, version-controllable, and portable to any tool or model you switch to next year.
When I built the AI operating layer for our CS organization, the highest-leverage decision wasn't prompt engineering or model selection. It was putting everything in files with clear names in predictable places. Customer context. Product grids. Reporting templates. Once the information had a stable, readable shape, the AI could navigate it reliably — and so could I.
That auditability isn't a nice-to-have. In a business context, it's the difference between something you can trust and something you can only hope works.
There's a secondary benefit worth naming: writing a markdown file forces you to articulate what you actually know. Agents let you skip that step. You describe the goal, wire up the tools, and let the model figure out the rest. The problem is that "the rest" often includes context that exists only in your head — and the model has to hallucinate a substitute.
The teams that got the most out of AI weren't the ones with the most sophisticated pipelines. They were the ones who had done the unglamorous work of writing things down. The file is the forcing function. The organization is a byproduct.
I'm not arguing against agents categorically. They're the right tool when a task requires dynamic decision-making across genuinely unpredictable inputs, when volume makes human-in-the-loop impractical, or when you need live integration across multiple real-time systems.
But that's a narrower category than most teams think. Before you architect an agent, ask: is this a coordination problem, or is it an information problem wearing coordination's clothes? In my experience, it's usually the latter. And the fix is a well-named file in the right folder.
The most powerful AI setup I've seen in production isn't a 12-step agent pipeline. It's a team that writes things down, keeps them current, and stores them somewhere predictable. The AI just reads along.